Results 1 - 10 of 52,676

Table 9: Multi-hop Ad-Hoc Network Model with Energy-Efficient Routing Link Met- rics

in Energy-Efficient MAC in Ad Hoc Networks Inspired by Conflict Resolution Concepts
by Yalin Evren Sagduyu, Anthony Ephremides

Table 1: Task cycle and energy efficiencies on the allocated hard- ware.

in System-Level Synthesis of Low-Power Hard Real-Time Systems
by Darko Kirovski, Miodrag Potkonjak 1997
"... In PAGE 1: ...n a particular processing element fPEj, j=1..3g. The Ci;j data are shown on the left side of Table1 . Each unit PEj consumes a specific amount of energy Ej per operation.... In PAGE 1: ... Thus, the energy PEi;j required for one execution of each task Ti can be computed as Pi;j=PEj Ci;j. The right side of the Table1 provides the data on the task execution energies Pi;j. The tasks are independent and have to be executed exactly once per each iteration (21 s).... ..."
Cited by 34

Table 4.2: Recognition results for high accuracy

in On Supervised Learning From Sequential Data With Applications For Speech Recognition
by Michael Schuster

Table 1: Task cycle and energy efficiencies on the allocated hard- ware.

in System-Level Synthesis of Low-Power Hard Real-Time Systems
by Darko Kirovski, Miodrag Potkonjak
"... In PAGE 1: ... j=1..3g.TheC i;j data are shown on the left side of Table1 . Each unit PE j consumes a specific amount of energy E j per operation.... In PAGE 1: ... Thus, the energy PE i;j required for one execution of each task T i can be computed as P i;j =PE j #01C i;j . The right side of the Table1 provides the data on the task execution energies P i;j . The tasks are independent and have to be executed exactly once per each iteration (21#16s).... ..."

Table 1: NTPC consistently yields improvements on all eight different high-accuracy NER base models, across every combination of task and language.

in Why nitpicking works: Evidence for Occam’s razor in error correctors
by Dekai Wu, Grace Ngai, Marine Carpuat 2004
"... In PAGE 3: ...MH as the base learner, which showed that NTPC was ca- pable of robustly and consistently improving upon the accuracy of the already-highly-accurate boosting model; correcting the errors committed by the base model but not introducing any of its own. Table1 compares results obtained with the base Ad- aboost.MH model (Schapire and Singer, 2000) and the NTPC-enhanced model for a total of eight different named-entity recognition (NER) models.... ..."
Cited by 1

Table 1: Training results with 3 output-units The gure 5 correspond to the neural network with 15 output-units and the gaussian output- representation (see section 4). The performance of this neural network is encouraging especially if you consider that the average error less than 10 degrees on the training data and 12 on the test data. There is some inaccuracy in the training data, since the gaze in a certain direction is only partially correlated with orientation of the head. Considering this we can conclude that the Gaz- etracker is able to determine the head-direction with a high accuracy.

in Gaze Tracking Based on Face-Color
by Bernt Schiele, Alex Waibel 1995
"... In PAGE 5: ... But the number of iterations was been always within the range 100{150. Table1 and gure 5 show the training- and test-results of the neural networks described in the previous section. Table 1 corresponds to the neural network with 3 output-units and the 1 from 3 representation (see section 4).... In PAGE 5: ... Table 1 and gure 5 show the training- and test-results of the neural networks described in the previous section. Table1 corresponds to the neural network with 3 output-units and the 1 from 3 representation (see section 4). This table shows the capability of the network architecture to distinguish between... ..."
Cited by 78

Table 1: NTPC consistently yields further F-measure gains on all eight different high-accuracy NER base models, across every combination of task and language.

in N-fold Templated Piped Correction
by Dekai Wu Grace, Grace Ngai, Marine Carpuat
"... In PAGE 3: ... We aim to further improve performance, and propose using a piped error corrector. 4 Results Table1 presents the results of the boosting-only base model versus the NTPC-enhanced models on the eight different named-entity recognition models, using differ- ent tasks and languages. For each task/language com- bination, the top row shows the base model (AdaBoost) result, and the bottom row shows the result of the piped system.... ..."

Table 1. Recognition results for high accuracy, cleaned of errors that shouldn apos;t be counted in

in Nozomi - A Fast, Memory-Efficient Stack Decoder For Lvcsr
by Mike Schuster 1998
Cited by 1

Table 2. Summary of Proposed Energy-Efficiency Measures

in Ukraine: Emerging Market for Industrial Energy Efficiency Opportunities
by Sriram Somasundaram Steve, Steve Parker, Meredydd Evans, Daryl Brown
"... In PAGE 6: ... The end-use energy efficiency measures are thus an integral part of the proposed strategy for improving the energy efficiency at the plant. Table2 lists the proposed energy-efficiency measures recommended for implementation at the Avdeevka coke-chemical plant. These measures are all cost-effective, with the... ..."

Table 2 lists the proposed energy-efficiency measures recommended for implementation

in Ukraine: Emerging Market for Industrial Energy Efficiency Opportunities
by Sriram Somasundaram Steve, Steve Parker, Meredydd Evans, Daryl Brown
"... In PAGE 7: ...Table2 . Summary of Proposed Energy-Efficiency Measures ( Avdeevka Coke-Chemical Plant) Measure Cost (US $) Annual Savings (US $) Simple Payback (years) Internal Rate of Return (IRR) (%/yr.... ..."
Results 1 - 10 of 52,676